Lead scoring is a methodology that assigns numerical scores to leads or accounts based on fit attributes and behavioral signals, enabling sales and marketing teams to prioritize outreach toward the contacts most likely to convert.
How Lead Scoring Works
Lead scoring combines two categories of data. Fit data captures whether a contact or account matches the ideal customer profile: industry, company size, revenue, technology stack, geography. Behavioral data captures whether the contact or account is showing purchase intent: pages visited, content downloaded, email engagement, ad interactions, and third-party intent signals.
Scores are assigned by weighting these inputs based on their historical correlation with conversion. A contact at a 500-person SaaS company who visits the pricing page twice and downloads a competitor comparison earns a higher score than a contact at a 10-person startup who opens one email.
Explicit vs. Predictive Scoring
Explicit scoring uses rules written by a human: "+10 for VP title, +15 for pricing page visit, -20 for student email domain." These are transparent and easy to audit but require manual maintenance as conversion patterns shift.
Predictive scoring uses machine learning to identify patterns in historical win data and apply them forward. Predictive models can surface non-obvious correlations but are harder to audit and can encode historical bias if the training data is limited.
Most modern teams use a hybrid approach: machine-generated base scores with human-maintained threshold rules for suppression and priority flags.
Example
A marketing automation team sets a threshold of 80 points before passing a lead to sales. A contact accrues 35 points for fit (target industry, company size, director-level title) and 50 points for behavior (three website sessions, pricing page visit, content download) and crosses the threshold after two weeks. The lead is routed to an account executive with the full engagement history attached.
How Abmatic Does This
Abmatic identifies accounts visiting your website without a form fill and layers in firmographic fit data, giving your scoring model account-level behavioral signals that anonymous traffic would otherwise miss.
Related: Intent data definition | Ideal customer profile definition
